Thermoelectric (TE) materials have unique advantages in directly converting any level of thermal energy into electrical power and solid-state cooling by a reverse mode. Although thermoelectric devices are regarded advantageously with their high reliability, their lack of moving parts, and their ability to scale to any sizes; the devices energy conversion efficiency remains generally poor. To improve the TE performance, many approaches have been investigated over a half century. The efforts that have pushed the figure of merit (ZT) of TE materials up beyond 2 have been reported by several groups by employing low dimensional nanostructures, such as nanowires and nanoparticles. However, the TE performances reported by most of researches were based on those laboratory developed materials and test results. The device level performance of those TE materials reported has fallen short of the reported level of TE materials. Consequently, the performance of practical TE devices using those reported TE materials have rarely been better than the figure of merit 1 due to junction resistivity and an apparent mismatch of Seebeck, coefficients of TE and junction materials.